> > Program received signal SIGSEGV, Segmentation fault.
> > get_exportlist () at mountd.c:346
> > 346 n = c->gr_next;
> >
>
> I guess 'c' is NULL at this point. Please ignore my last patch and
> try this one instead.
HJ,
I tried it - no joy. The symptoms are even more bizarre:
----------------------------------------------------------------
Here's what happens when I mount from the alpha, then run 'showmount -e'
on the server:
Script started on Mon Oct 12 19:41:57 1998
(root@cy) /usr/src/net/nfs/knfsd-981010/utils/mountd : showmount -e
Export list for cy:
/ *.steve.net,*.fast.net
/usr/src *.steve.net,*.fast.net
(root@cy) /usr/src/net/nfs/knfsd-981010/utils/mountd : gdb mountd
GDB is free software and you are welcome to distribute copies of it
under certain conditions; type "show copying" to see the conditions.
There is absolutely no warranty for GDB; type "show warranty" for details.
GDB 4.16 (i586-unknown-linux), Copyright 1996 Free Software Foundation, Inc...
(gdb) att 1550
Attaching to program `/usr/src/net/nfs/knfsd-981010/utils/mountd/mountd', process 1550
Reading symbols from /lib/libc.so.5...done.
Reading symbols from /lib/ld-linux.so.1...done.
0x40079b3c in __select (nd=-1073742508, in=0xbffffd64, out=0x0, ex=0x0, tv=0x0)
(gdb) cont
Continuing.
Program received signal SIGSEGV, Segmentation fault.
0x4008cb7e in libc_nls_init ()
(gdb) bt
#0 0x4008cb7e in libc_nls_init ()
#1 0x8054960 in ?? ()
(gdb) kill
Kill the program being debugged? (y or n) y
(gdb) quit
(root@cy) /usr/src/net/nfs/knfsd-981010/utils/mountd : exit
Script done on Mon Oct 12 19:44:03 1998
------------------------------------------------------------
And here's what happens when I mount from the alpha and run 'showmount -e'
from the client:
Script started on Mon Oct 12 19:46:45 1998
(root@cy) /usr/src/net/nfs/knfsd-981010/utils/mountd : showmount -e
Export list for cy:
/ *.steve.net,*.fast.net
/usr/src *.steve.net,*.fast.net
(root@cy) /usr/src/net/nfs/knfsd-981010/utils/mountd : showmount
Hosts on cy:
(root@cy) /usr/src/net/nfs/knfsd-981010/utils/mountd : gdb mountd
GDB is free software and you are welcome to distribute copies of it
under certain conditions; type "show copying" to see the conditions.
There is absolutely no warranty for GDB; type "show warranty" for details.
GDB 4.16 (i586-unknown-linux), Copyright 1996 Free Software Foundation, Inc...
(gdb) att 1719
Attaching to program `/usr/src/net/nfs/knfsd-981010/utils/mountd/mountd', process 1719
Reading symbols from /lib/libc.so.5...done.
Reading symbols from /lib/ld-linux.so.1...done.
0x40079b3c in __select (nd=-1073742508, in=0xbffffd64, out=0x0, ex=0x0, tv=0x0)
(gdb) cont
Continuing.
Program received signal SIGPIPE, Broken pipe.
0x40078968 in __write ()
(gdb) bt
#0 0x40078968 in __write ()
#1 0x4005d7ae in writetcp (xprt=Cannot access memory at address 0xfb4.
)
Cannot access memory at address 0xfa0.
(gdb) kill
Kill the program being debugged? (y or n) y
(gdb) quit
(root@cy) /usr/src/net/nfs/knfsd-981010/utils/mountd : exit
Script done on Mon Oct 12 19:49:21 1998
------------------------------------------------------------------
Some more information:
The problem is not particular to the Alpha as client. I fired up an older
386 machine that sits on 'steve.net' (192.168.244.0) along with the Alpha.
Same problem crops up. Boxes on the other interface, 'fast.net'
(192.168.245.0) work fine.
I originally thought the problem existed solely on the router, but another
box on fast.net (Dual PPro) blows up when mounted from the other net...
To make matters worse, my trusty 486 box that sits on fast.net and carries
the cable-modem interface has no problems at all - even with cross-subnet
mounts! I can't see any common denominator between the two servers with
mountd problems - nor anything that sets them apart from the 486.
FWIW, 'steve.net' interfaces through eth1 (10Base-T Tulip PCI adapter) and
'fast.net' through eth0 (100BaseTx Tulip PCI adapter). These problems
started with kernel 2.1.125 & knfsd-981010.
Another observation: The mount daemon now ignores the 'insecure' export
option. In particular, the Berkeley automounter (amd - used on all my
BSD* boxen) no longer works reliably. It tries to use port 1023, IIRC.
So, something broke here as well.
Keep sending me patches and I'll keep pounding on it!
Steve
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/